Responsibilities Have Shifted

It used to be fairly universal practice to move the texts of the various languages directly from the translation source(s) to the graphic design studio, print production house, or commercial prepress firm responsible for getting the material from manuscript to press-ready form. But in today’s era of "desktop publishing, translators themselves, or staff at translation companies may be called upon to handle the page layout function and create files ready for the print shop. Either way, it becomes important for translators, translation project managers, and/or end-user client personnel to understand and appreciate the nature of problems presented by multilingual requirements and the range of available solutions.

A given project may involve just one or two, or as many as several dozen different languages. I have over the years worked on projects in which the same publication or informational text needed to be produced in 30 or more different languages. When it is important to present an attractive appearance, provide good functionality, meet legal informational mandates (e.g. food packaging), and hold production costs to a reasonable level, the know-how of experienced professionals in preparing the material is almost invariably required. And the pros are often confronted with some rather daunting aesthetic and physical problems.

Perhaps foremost among the problems that bedevil the designer of a multilanguage project is the fact that languages take different amounts of space to say the same thing. On the printed page, space has finite limits, and the variation in text length among languages must be dealt with in a satisfactory manner, lest the final product be an ugly hodge-podge, unattractive and difficult to read or use.

The variance can be - and often is - so great that an easy adjustment in page format is not sufficient to make the accommodation. This is particularly true from the point of an American or British typographer because English tends to be more "economical" with respect to text length than any of its Western counterparts. If the original design was planned for English text without provision made for subsequent employment of other tongues, the hapless typographer finds himself overly restricted in the choice of design solutions available.

No Infallible Formulae

In any event, though, there will be a length variation among the languages involved, and this phenomenon must be accounted for in the final layout of the printed pages. The variation can be rather large, although there are no infallible rules of thumb for guidance.

It is true that individual style and "wordiness" of a particular translator can noticeably affect the length of a translation; but, it is a rare exception with literary or commercial expository texts to find a translation from English into a Western language that is not significantly longer than its original.

An interesting and instructive illustration of this fact of linguistic life can be found in the publication by the U.S. Library of Congress in 1972 of a booklet in which are presented translations by selected world literary figures of Lincoln's Gettysburg Address.

Every attempt was made - more difficult in those days of metal typesetting - to maintain a uniform size and appearance of the various settings. While the sponsor's intent in seeking this uniformity of typographic style was aesthetic, a comparison of the resultant lengths in print is instructive. The Greek translation, for example, ran some 24% longer than the English original; Hungarian occupied 28% more space; and German 32%.

Typographia polyglotta

More recently, a deliberate attempt to study and quantify this phenomenon was made by Prof. George Sadek and U.N. Graphics Chief, Maxim Zhukov, as a research project of the Center for Design and Typography of New York's Cooper Union, the results of which were published in a book entitled "Typography:Polyglot" in 1991. The publication was amended, updated, reformatted, and republished in 1997 by ATypI (Association Typographique Internationale), retitled as "Typographia polyglotta," in connection with its annual meeting held that year in Reading, England.

Sadek and Zhukhov took as their "control" a passage from the preamble to the United Nations Universal Declaration of Human Rights. They then had my firm typeset the official U.N. translation of the selected text in 30 other languages, attempting in all cases to maintain as far as possible (inasmuch as a number of non-Latin scripts were involved) a uniformity in the typography which would permit a useful comparison of the physical "area" required to print each language.

As had been the case with the Gettysburg Address booklet, results of the "Typography:Polyglot" study showed a considerable variation in the space required for each language compared to the English "control." Czech occupied 17% more space; Dutch required a 28% greater area; and Greek turned out to be the lengthiest, occupying 129% of the area needed for the English!

Actual Experiences Vary

My own experience, working mostly with advertising copy, public relations materials, and marketing communications of various kinds, has differed somewhat from the cases described above. In the majority, I've noted that (again, allowing for translators' individual idiosyncrasies) all European languages require significantly more space in translation than does an English original, ranging from an average of about 20% longer for Spanish to as much as a 50% increase for German or Russian. The length issue is exacerbated for Russian due to the nature of its Cyrillic script, which has no "narrow" characters (e.g. the Latin "i" or "l").

I did find, however, that some languages, notably Chinese, run considerably shorter than English; and that many, such as Arabic, Hebrew, Japanese, and Korean, almost always will fit with equivalent legibility within the same space or less than that needed for English. Nevertheless, in the present article, because of space limitations, I will confine discussion of copyfitting/layout problems caused by variations in text lengths to the Western languages printed with the Latin, Cyrillic, and Greek scripts, inasmuch as the design solutions presented are common to all of them - which is not necessarily true of languages reading in different directions, employing connected scripts, or lacking the distinctions of upper and lower cases.

Four Paths to Follow

Even within this limited language range, workable solutions must be found to the inevitable difficulties caused by the variance in space required by each of the languages involved. Naturally, the approach taken will be partially dependent on the number of languages involved and the nature of the project; but, overall, the approaches usually found best tend to fall into four broad categories, which I shall characterize as: Spatiality, Separation, Sequencing, and Synthesis.

In an optimal scenario , the multilanguage requirements of a project are taken into account right at inception, and allowances made accordingly in the original layout plans. It is especially helpful if all translations are completed before an overall design is finalized, so that a typographic strategy can be formulated to best accommodate the actual texts.

In the real world, however, this is most often not the case. One of the worst - and not uncommon - situations in which my composing room has found itself is to be presented with a printed English document, set in small size type (or, worse, a condensed face), filled up with text right to the margins on many or most pages, and then be asked to typeset versions in French or Russian or German within the same copy areas. Usually, when this happens, it simply cannot be done with any semblance of maintaining legibility and/or a satisfactory visual aesthetic.

An Unsatisfactory Arsenal

The typographer's arsenal in dealing with such situations includes the weaponry of reducing type size, condensation of the fonts used, expanding text further into margin and "white-space" areas, requesting copy cuts from client or translator, minimizing interline spacing, and whatever else one can think of in an attempt to fit 10 kilos of copy into a space designed for five. The results are invariably less than good.

As I have indicated above, when there is an opportunity - and the motivation - for advance planning, a number of approaches to solving the problems of varied text lengths may be applied, singly or in combination, depending on the specifics of any given problem; and no one solution fits all. For purposes of this discussion, let us consider the most commonly used methods referred to above.

Spatiality: Linguistic Elbow Room

The first of these approaches simply requires designing for the longest language of the project. That is, the layout's creator should allow sufficient space for the longest of the texts to be set in a legible type size with suitable interline spacing. If this is done, it follows that all shorter texts will be easily accommodated, leaving blank space where they fall short of the original designed area.

This presupposes that the longest of the texts will be available for copyfitting at the time of design, though this is not always the case. Often, the graphic designer has only a single manuscript on which to base the layout; and it may not be the shortest of the languages that will eventually be used. This is very common in the US, where designs are often created from the English original well before any translation has even begun.

The principle, nevertheless, can still be effectively utilized by projecting (based on rough rule-of-thumb guides and/or prior experience) approximately how much space will eventually be required for the longest language, and then using that as a basis for the English version. The concept of designing to accommodate the longest language text is also useful when employed in combination with the other techniques discussed below.

Separation and Sequencing

Separation of each language version into a discrete, separately published version, is a costly but highly effective method to maintain a uniform look even when the languages required vary not only in length, but also in kind of script used and direction of reading.

This approach is particularly suited for larger projects or publications which will be distributed separately to different audiences; and I have often worked with clients to produce excellent results using this approach in creating multilanguage series of such publications as company policy manuals, employee benefits handbooks, utilities' information pamphlets, and similar materials.

Some multilanguage materials are printed in sequential form. That is, within a single printed piece, each language is printed in its entirety in sequence, one after another. Each runs for as long as is necessary to accommodate the length of its text; the next language follows; then the next; and so on until all have been presented.

This strategy is particularly effective when either a small number of languages is involved, or the text is quite short. In the former case, as exemplified by the instruction books accompanying my VCR, a camera, and a tape deck, only two or three languages are presented. The reader can comfortably refer to the language in which he or she is fluent, and the overall booklet does not become unwieldy in size.

When texts are short, many languages can be accommodated easily in a sequential manner. This is often seen in such publications as the multipane fold-out info sheets that come with film rolls. We used to translate and typeset very successfully in this format instructional package inserts for surgical supplies in as many as thirteen different languages.

One drawback of the sequential language layout is that photos and diagrams, if any, must be repeated with each of the language texts. If the graphic elements are too numerous, this can significantly increase the publication's size and/or the printing costs.

Synthesis

The first definition in the new American Heritage Dictionary (4th ed.) for synthesize is "To combine so as to form a new, complex product," which describes another design approach to multilanguage printing fairly well. When the languages are relatively few in number - and especially when all are printed in European scripts - it is often feasible and desirable to print them together on each individual page.

Two, three, or as many as four side-by-side columns may be utilized on a page. Each can have a different vertical length, depending on the particular language's text. All should be aligned at the top, with the shorter languages leaving more blank space below their columnar areas.

Similarly, as many as five or six different languages, dependent on text length, can be presented as horizontal bands on each page. The depths of the bands can vary to account for differing text lengths. If the languages are more numerous, or the lines run too long for easy legibility, individual complete copy blocks can be presented in two side-by-side columns; again allowing the depth of each block to vary according to need.

This intermixing principle can also be used for short diagram or picture captions, with each language set, one under another, beneath the referenced illustration.

The Virtues of Experience and Timing

In all such cases, when intermixing languages on each page, it is essential that the same language be located in the same relative position on every page; that (except for short captions) each text be of reasonable length; and that all text blocks be placed on the same page or spread with any illustrations to which they refer. Each reader may be expected to read or search for information only in his or her own tongue, and it should be easy for that reader to locate the wanted language every place it appears. In some layouts, it is possible (and usually desirable whenever graphically possible) to tag with symbols and/or color-coding each language for the reader's convenience in finding the sections in the tongue of his choice.

In the world of print communications, multilanguage projects - whether involving two or two dozen languages of the old - present special challenges in legibility, utility, and aesthetics. In every case, the advantage lies with the experienced graphic designer and the project manager who can and will plan ahead, beginning with the earliest possible point in a project's conception and production.

---

Richard N. Weltz
Marketing & Business Management Consultant
360 East 72nd Street, New York, NY 10021, USA
Voice: +1 212 734 6285 Fax: +1 212 879-0287
Email: Rweltz@att.net

---

Richard Weltz has worked with typesetting from hot metal through today's PostScript. He founded and presided for the past 35 years over Spectrum Multilanguage Communications, a translation firm with uniquely extensive capabilities in multilanguage typography, and is now an independent consultant to the language industry.